Generating Synthetic Free-text Medical Records with Low Re-identification Risk using Masked Language Modeling

TL;DR

This paper introduces a Masked Language Modeling system for generating synthetic medical records that balance data utility with privacy, achieving high-quality data with low re-identification risk and cost-effective inference.

Contribution

It presents a novel Masked Language Modeling approach for synthetic medical data that preserves privacy while maintaining data diversity and utility.

Findings

High-quality synthetic data with 96% HIPAA-compliant PHI recall

Re-identification risk reduced to 3.5%

Generated data enables effective model training comparable to real data

Abstract

The vast amount of available medical records has the potential to improve healthcare and biomedical research. However, privacy restrictions make these data accessible for internal use only. Recent works have addressed this problem by generating synthetic data using Causal Language Modeling. Unfortunately, by taking this approach, it is often impossible to guarantee patient privacy while offering the ability to control the diversity of generations without increasing the cost of generating such data. In contrast, we present a system for generating synthetic free-text medical records using Masked Language Modeling. The system preserves critical medical information while introducing diversity in the generations and minimising re-identification risk. The system's size is about 120M parameters, minimising inference cost. The results demonstrate high-quality synthetic data with a HIPAA-compliant PHI recall rate of 96% and a re-identification risk of 3.5%. Moreover, downstream evaluations show that the generated data can effectively train a model with performance comparable to real data.